Risk Acceptability According to the Social Sciences

7. Risk-Seeking and Safety First

Cltapter 7 Risk-Seeking and Safety "rst Summary: This chapter traces one line of development in the theory of choice that follows Simon's work on bounded rationality. He drew attention to the enormous intellectual complexity of decision-making assumed by the theory. By substituting "satisficing" for maximizing he has been able to synthesize many observations of how decision units actually behave, assuming that they accept upper and lower thresholds of failure and success. Research in agricultural economics indicates how these limits are culturally defined. Rational choice needs to take into account the working of a prinCiple that sets the lower limit for acceptable risks. Utility theory allowed for outcomes being so undesirable as to fall outside of the individual's schedule of preferences. Engel's law (Houthakker 1957) recognizes a kind of safety-first concern. According to this observation, the poorer the households in a given income distribution , the lower is the elasticity of response to changes in the price of food. The household head has an order of spending that strives to ensure the basic requirements of his family. In investment analysis, liqUidity preference is a similar idea, which, according to Hicks, entered economic discourse in the 1930s. An investment trust or discount house has liabilities that must be protected: "a worse than 'expected' outcome must be dreaded more than a better than expected outcome is desired-because of the impact which unfavorable outcomes may have on the nonliquid elements in the situation . . . hence the motive force for certaintv" (Hicks 1962). The snnplest example of this concern to limit the worst outcomes is the practice usual among pastoralist tribes to divide a 73 big herd of livestock among several dispersed herdsmen. It corresponds to Daniel Bernoulli's advice that it is advisable to divide goods which are exposed to some danger into several portions rather than to risk them all together (Bernoulli 1738: 30, par. 16). Early attempts to understand behavior under uncertainty used the maximizing of expected gain as the main criterion of rationality . But some lower limit of tolerable loss was also recognized (Roy 1952). The various bits and pieces of a general safety-first principle became formalized within Simon's concept of bounded rationality (1955). The ideas embodied in this paper were initially developed in 1952 (see Simon 1979:7), which puts it in the same short period as other famous questionings of rational choice theory (see chapter 4). In this decisive new development Simon starts with two reasons for questioning the adequacy of the theory of rational choice. One is the grotesquely powerful intellectual capacities which are supposedly called upon for every choice. The theory expects an impossibly complex analytic exercise from the rational agent. The other is the neglect of the rational agents' environment : "We must be prepared to accept the possibility that what we call the 'environment' may be, in part, within the skin of the biological organism." This is a useful approach to the fact that the individual's thinking apparatus, with its concepts and assessments of the world, though it may lie within his skin, is also part of his environment. "SatisfiCing" is Simon's word for the method of simplifying complex choices, which he argues is more commonly used than the "maximizing" behavior assumed by utility theory. A satisficing fiml sets its goals to the attaining of a certain level of profit, to holding a certain share of the market, or to keeping up a certain level of sales. Maximizing behavior does not put any limit on what may be achieved. SatisfiCing is adaptive behavior. In this approach, rational choice takes place within bounds whose upper and lower limits are set semi-independently. This is consistent with psychological studies of the formation and change of levels of aspiration. It is assumed that goals have some connection with what is attainable. In classical utility theory, the zero point in the scale of utility is set arbitrarily; in this approach, the adaptive aspiration level defines a zero pOint and a ceiling. Simon cites evidence that firms with a declining share of the market strive more Vigorously than firms whose share in the market is steady or in74 creasing. There is further evidence from peasant communities where a socially set standard of living defines upper and lower limits of striving (Sahlins 1974). This hypothesis has inspired much empirical research on risktaking . However, the clear implication that the levels of aspiration...